Credit: Particle-in-cell simulation of the wake generated in a lithium plasma by a 42 GeV electron beam.

The electric field within any given section of a conventional particle accelerator is limited by the breakdown field of the materials from which it is made. Consequently, getting to the relativistic energies where interesting particle physics takes place requires multiple accelerator stages spanning many kilometres. The extreme fields generated when a high-power laser is focused into a plasma have been suggested as a way of accelerating particles in a much smaller space, and potentially with less expense. On their own, accelerators that operate on this principle — known as plasma wakefield accelerators — can generate electron beams with energies exceeding 1 GeV, in a plasma just a few centimetres long. This is, of course, far below the energy demanded by particle physicists. But by using wakefield accelerator techniques in tandem with the conventional linear accelerator at the Stanford Linear Accelerator Centre, Ian Blumenfield and colleagues show they can use them to double the energy of electrons from a 42 GeV electron beam over a distance of less than a metre (Nature 445, 741–744; 2007).

Plasma wakefield accelerators typically operate by focusing a high-power laser pulse on a tight spot in some medium, which can be a gas, liquid or solid. This creates a fully ionized plasma and drives freed electrons through the plasma at close to the speed of light. It is the wake left behind by this burst of ultra-relativistic electrons that generates the large fields used for particle acceleration. But a high-power laser is not the only means to drive electrons through a plasma to produce such a wake, a fact that Blumenfield et al. clearly demonstrate by focusing their conventionally accelerated electron beam into an 85-cm-long column of lithium vapour. When the beam interacts with the vapour it transfers most of its energy to the resulting plasma, and the wake it produces accelerates a proportion of electrons in the tail of the beam to twice their original energy (see 'particle in cell' simulation pictured). Simulations of this interaction confirm an implied accelerating field of 53 GV m−1 — a thousand times greater than can be achieved in a typical linear accelerator stage.